Childhood ALL is the most predominant form of cancer in children, but its causes and risk factors remain largely unknown. A few factors, such as Down's Syndrome, other specific genetic abnormalities, being an identical twin of a sibling with leukemia, and ionizing radiation exposure (both in utero and postnatal), have been consistently linked to childhood ALL and acute myeloid leukemia (AML); however these conditions only explain a small fraction of cases (Robison L, 1995). Indeed a plausible etiologic explanation is lacking for more than 90% of cases of childhood leukemia (Pui C-H, 1995). In the March 2002, a workshop was convened at the National Cancer Institute to identify urgent research needs and facilitate collaborative interdisciplinary approaches (Linet M , 2003). During the NCI Workshop on Gene-Environment Interactions in the Etiology of Childhood Cancer, the idea of a pooled analysis of birth weight and childhood leukemia was born. In a presentation entitled "The application of the life cycle paradigm to childhood cancer epidemiology", I discussed the relation of high birth weight to childhood ALL. Using the review by Ross (1996) that indicated an overall 2-fold increased risk of leukemia in children of high (> 4,000 g) compared to normal birthweight in 12 of the 16 studies, I reviewed the research since 1996. My talk extended the association to all but one of the recent studies (Westergaard T, 1997; Ross J,1997; Roman E,1997; Reynolds P, 2002), identified the potential for a similar association for AML and the potential for a greater magnitude of effect of birth weight in children who were diagnosed with cancer aged < 5 years versus those aged > 5 years. My talk ended with a description of the major determinants of birth weight, (Hennessey E, 1998) and examined the potential role of early childhood growth in the etiological pathway to ALL based on the research of Westergaard (1997) and of Broomhall (1983), that is similar to work in breast cancer (De Stavola B, 2000). I called for researchers to examine the effects of the leading determinants of birthweight on risk of childhood leukemia. During the meeting, Dr. Joachim Schuz and others indicated their interest in developing a collaboration to examine the role of birthweight in childhood leukemia by diagnostic subtypes as well as by other factors, because childhood ALL and AML are rare diseases with too few patients in a single study to perform the analysis of interest. Within the year the idea for a pooled analysis of this area had been fleshed out by Dr. Schuz and I, and an invitational letter was sent to potential collaborators. AimsThe aims of the pooled analysis of birthweight and childhood ALL and AML are to:1. Examine the association of birthweight to childhood leukemia by age at diagnosis and by diagnostic subtypes;2. Should a birthweight association appear, examine the influence of the major factors associated with birthweight and whether infant feeding or growth in early childhood modifies the relation.